Year of Graduation
Geosciences data mining
School of Applied Mathematics and Information Science
Our world consists of different areas which contain a massive amount of data and interact with each other. Scientists apply computer-oriented operations of detecting patterns in voluminous assemblies of data comprehending techniques at the intercrossing of artificial intelligence, machine learning, statistics, mathematic and database systems. Basically, the overall problem is to pick out information from databases and transfashion it up to user-friendly organization. The geographic study outburst is not quite different from resembling revolutions in marketing, biology, and astronomy. Today the problems associated with the study of the geographic environment attract a strong interest in modern Data Analysis and Computer Science. Recently, newly developed methods of data processing and analysis are focused on the active participation of experts of the data domain. Therefore, the principal goal of this paper is to group data of the marine environment. With the help of the new method of the cluster analysis "Carticlus", I will reveal sea areas with similar trends (parameters) and draw an analogy between obtained information and reality. For our study purposes we use State Research Center "Arctic and Antarctic Reserch Institute" database of the Barents Sea parameters observations, which was composed of seven main parameters:temperaturetotal salt contentlatitudelongitudeAs a result I have realization of Carticlus algorithm and results of clasterization of our data.